Sometimes a business problem that sounds simple to solve turns out to be a major challenge. For example, getting a count of a company’s customers sounds easy. But within an enterprise, different information systems may have slightly different data for the same customer, producing a new, duplicate record. The same customer can get counted twice, making the total number of customers incorrect. A company also needs to determine their definition of a customer and get agreement across business units and functional areas. Sales may define it at the account level, while operations defines it at the location or operating unit level. The same holds true for product records--a product may be listed in different ways, so analyses that rely on connecting a metric with a product may be inaccurate.
When a company is working on a business intelligence or an analytics project, data quality is often a roadblock. Analytics that are run on poor quality data will produce results that are not in line with expectations, or do not pass the common sense test. Inconsistencies make aggregating data or analyzing it across different functional areas nearly impossible. Millions of dollars can be spent without the project having achieving the desired outcome, such as gaining insights into businesses operations or tailoring promotions to a particular customer segment.
Another issue that can impede effective analyses is that employees want to solve business problems but they do not understand the data they have. Often, the perceived solution is the deployment of yet another software product. But this can create yet another set of data that may be inconsistent with other systems. Or, data may be updated in one system but not in another, creating inconsistencies across the enterprise. The core issue is not the capabilities of the software so much as it is the issue of data governance and the quality of the data.
The solution is to ensure that there is one set of accurate data or golden record, and the way to do that is through Master Data Management (MDM). An enterprise-wide information architecture that supports the linking of master data to enterprise data is a mechanism for understanding business concepts and then transferring that into data elements. Ideally, the information architecture will capture the business landscape, and the exercise of developing it moves a company shifts the perspective from diverse data sources to a holistic view of enterprise information.
EIS provides unique value in having an end-to-end understanding of business technology. Many companies can establish their business intelligence (BI) systems or implement MDM, but do not have a view of the information flow from its origin to how it gets transferred from data to tangible, actionable information. Systems integrators attempting to assist them may understand the technology, but cannot follow the information flow needed to complete business activities.
Companies generally realize when they have an issue that needs to be solved by an MDM project. They have inconsistent data that is causing a problem, or are unable to determine whether an action on one element in an organization will impact another. For example, when a transaction takes place with one customer, what else is affected? Is a customer also a supplier? or a potential partner? The lack of clear relationships among data element impedes achievement of business goals.
In order to answer questions like this, a robust data model that organizes the data elements and shows the relationships among them must be developed. A wireframe that shows what the user experience will look like, where data is surfaced in different places and what the search facets are, is a useful tool. But ultimately, data quality is fundamental to customer experience, along with having master data and integrating it into critical business applications.
For a deeper understanding of the important role of taxonomy in your MDM program read our white paper: Launching a Master Data Management Program: Eight Key Steps in the Journey